AI Policy
Annali Italiani di Chirurgia endorses the ICMJE recommendations on the use of artificial intelligence–assisted technologies (e.g., large language models [LLMs], chatbots, or image generators) and requires that all authors, editors, and reviewers adhere to these guidelines, ensuring the responsible use and transparent disclosure of AI involvement.
Authors
Authors should note that artificial intelligence–assisted technologies do not meet the criteria for authorship and therefore must not be listed as authors or co-authors.
Any use of traditional or generative AI tools in creation, review, revision, or editing for manuscript preparation must be transparently disclosed. Authors are required to specify the name and version of the AI tool(s) and describe in detail how they were used in the Declaration of Generative AI and AI-Assisted Technologies in Manuscript Preparation section. The disclosure does not extend to basic tools used for grammar, spelling, or reference checking.
Authors should recognize that AI can generate outputs that seem authoritative yet may be inaccurate, incomplete, or biased. Therefore, all AI-generated content requires careful verification to prevent potential errors. Authors are fully responsible for the accuracy, integrity, and originality of their manuscripts, alongside any materials generated with the assistance of AI. Editors reserve the right to make the final determination regarding the appropriateness and acceptability of AI use in submitted manuscripts.
Editors and Peer Reviewers
Annali Italiani di Chirurgia employs a double-blind peer review process. Manuscript information and all details related to peer review must be treated as strictly confidential. Editors and reviewers should recognize that the use of software or AI technologies poses potential confidentiality risks, as these tools may retain any provided prompts or manuscript content. Uploading unpublished manuscripts to software or AI tools that cannot guarantee confidentiality constitutes a breach of submission confidentiality.
Editors must not employ generative or AI-assisted technologies in the manuscript evaluation or decision-making process, as these systems are incapable of the critical thinking and independent judgment required for such work.
Similarly, reviewers should not use AI to conduct assessments of a manuscript. However, they may use AI tools as an aid to enhance the language, logic, or readability of their review reports. Any such use must be disclosed within their reports.
Both editors and reviewers should recognize that AI-generated outputs may appear authoritative but can be inaccurate, incomplete, or biased. Accordingly, all AI-assisted content must undergo careful verification to avoid potential errors.